This article was originally published on April 22nd, 2022. It was republished here without modification after the wallet address for shuel.eth was changed.

“The Web as I envisaged it, we have not seen it yet. The future is still so much bigger than the past.” -Tim Berners-Lee

The market is staggeringly saturated with smart contract platforms attempting to address the blockchain trilemma. 29 smart contract platforms currently command a market capitalization greater than $1,000,000,000. Dozens upon dozens of additional platforms have failed to achieve or receded below the $1B market capitalization line. During the next 3-4 years, I expect Solana to outperform all other existing non-Ethereum smart contract platforms. Key metrics by which to quantify this outperformance include, but are not limited to: daily active account growth %, market capitalization % increase, and total value locked (TVL) growth %. Solana will outperform its competitors because its design is distinctly different from the eerily similar design choices overlapping most of the market. Let’s examine what separates Solana from the rest of the market in detail.

Nearly all competitive smart contract platforms share a design choice: modularity. Modularity is generally defined as the decoupling of one or more of the consensus, execution, and data availability layers. The modular approach attempts to solve scalability bottlenecks that have traditionally plagued Ethereum. A review of the three layers and their respective bottlenecks can be found here. Modularity is juxtaposed with monolithic design, in which the three layers remain co-located. The general definition for modularity isn’t quite broad enough, however. A few examples will demonstrate why:

Example 1: Ethereum’s roll-ups are modular because they only perform execution; consensus and data availability are retained by the main Ethereum chain. The main Ethereum chain, however, still handles execution, consensus, and data-availability. In this sense, Ethereum is monolithic and only its roll-ups are modular. Polkadot employs a slightly varied design: parachains handle execution, while the relay chain handles consensus and data availability. The technical design differences are nuanced, complex, and outside the scope of this article.

Example 2: The Cosmos blockchain attempts to achieve scalability by acting as a “blockchain of blockchains” in which many independent blockchains are interconnected via the inter-blockchain communication (IBC) protocol. In this design, scalability is achieved by building a network of fragmented, monolithic blockchains that each handle their own consensus, execution, and data availability. Avalanche’s subnets are of a similar overall design: a network of fragmented, monolithic blockchains. This design is commonly referenced as being modular in nature.

Example 3: Ethereum 2.0, Zilliqa, and Near all employ sharding techniques, in which the data layer is spread out horizontally in order to improve scalability. This technique is modular in nature, as data becomes separated from consensus and execution.

Given these examples, modularity’s common definition is far too narrow. It may be more appropriately defined as any scaling solution that is not completely monolithic i.e. any design in which execution, data, and consensus for the entire network are not all handled in the same place.

Modularity improves scaling, but replaces it with another problem: composability is hindered. Composability is the capacity for system components to be recombined into larger structures and for the output of one component to be the input of another. This is commonly referred to as “money legos.”

Multicoin Capital’s Kyle Samani provides a succinct explanation of composability and its impacts on DeFi below:

Composability enables the development of higher order decentralized finance applications that have never previously existed. For example, the UXD protocol is a novel approach to algorithmic stablecoins using Solana DEXs (Mango currently) to fully collateralize issued stablecoins 1:1 with a delta-neutral position using derivatives. These protocols are easily composable on Solana because data, execution, and consensus for both happens in the same place.

Modular designs hinder composability for several reasons:

1. Syntactic composability is the ability of dApps to integrate and interact with one another. In avalanche, as an example, it is hindered (a) between the main network and its subnets and (b) subnet to subnet. Imagine that the UXD protocol exists on one subnet and Mango exists on another subnet. Every protocol interaction must be communicated across the two subnets. Now imagine a not-yet-developed protocol that requires interaction with a dozen other protocols spread across a dozen different subnets in order to generate its desired output. Bridging messages and tokens in that scenario would utilize excessive time that is unacceptable for many financial applications and increases failure risk. While Avalanche is developing solutions for rapid bridging and communications between subnets, monolithic designs will continue to retain a significant composability edge for the foreseeable future.

2. Atomic composability, the ability to bundle multiple operations into a single transaction, is impacted in a similar manner by modularization. Assets may become stranded or lost if one operation in the transaction succeeds while another fails. This is a risk commonly discussed regarding ETH 2.0’s sharding. Solana retains complete atomic composability due to its monolithic design.

No other non-Ethereum blockchain seriously competing to become a global settlement layer employs a fully monolithic design. This unique property will serve Solana well over the next several years. Solana developers will continue to produce novel DeFi products like UXD while competitor blockchains struggle to develop complex solutions to their composability issues.

The potential of composability cannot be overstated. Some of the coming financial creations will be truly revolutionary. Solana has over 300 DeFi applications currently running or under development. As developers create novel financial products, it will attract more developers, users, and TVL. This will result in a positive-feedback loop as the network’s growth leads to the incubation of additional applications and infrastructure on Solana.

Solana currently executes slightly under 3,000 transactions per second (tps) with block times of 400 milliseconds. It boasts a theoretical capacity of 65,000 tps. Furthermore, these transactions cost fractions of a penny - an average of $0.00025 / transaction. Founder Anatoly Yakovenko claims eight distinct innovations make this throughput and cost/tx possible. You can read more about the technical side of Solana here. Sub-second block times are necessary to attract traditional financial institutions that require rapid settlement times that aren’t tenable on chains like Ethereum.

Solana still advertises its mainnet as a beta product, despite its $7.36B TVL. Its reliability reflects this as well, with less-than-rare periods of severely degraded performance and infrequent complete outages. On September 14th 2021, the network went down completely for 17 hours during a denial-of-service attack.

https://twitter.com/SBF_FTX/status/1485839807708385283

Investors and Solana core developers take a collectively unconcerned approach to periods of degraded performance because (1) Solana still performs more TPS than most other blockchains while degraded and (2) it’s still advertised as a beta product under development. The developers are transparent and responsive in pushing code updates to the validators after each incident.

When compared against other Layer 1s, Solana appears most ready for developers to build applications for the masses despite its periods of degraded performance. There is no scaling solution coming at some arbitrary point in the future. The protocol is (mostly) ready to scale today.

Along with its monolithic design, Solana’s EVM non-compatibility is fairly unique. This has significant implications on the retention of resources within Solana’s network.

In EVM-compatible networks like Avalanche, protocols from Ethereum can easily be forked and deployed. Developers quickly fork previously successful protocols onto a new blockchain and collect available grant/incentive funds from that blockchain. These funds are then used to incentivize users to migrate. This historically produces cycles of mercenary capital rotation and vampire attacks, as investors and developers flow to the newest EVM-compatible blockchain with the best incentives. They then quickly move on to the next nearly-identical ecosystem once available incentives have been drained or a new blockchain offers slightly better subsidized incentives.

Solana utilizes the Rust programming language, which is substantially different from Solidity. This has several implications on retention of capital, users, and protocols:

1. Unique, successful applications built on Solana can’t easily be forked onto other blockchains. If individuals want to utilize the application, they must on-board into the Solana ecosystem. Bridging between Solana and EVM chains was originally difficult, further increasing the retention of capital and users on Solana. Bridging to and from Solana is rapidly becoming simplified and no longer poses a significant challenge to users, though.

2. Developers who dedicate time to building proficiency in Rust on Solana are far less likely to leave the ecosystem vs. Solidity developers who can jump to the next EVM chain with better incentives. With over $2,000,000,000 invested into the Solana ecosystem, I expect a magnitude increase in the number of protocols, developers, and users on Solana in 3-4 years. Those developers will be far more likely to stay working on Solana long-term compared to their peers in other ecosystems.

3. The previous discussion on composability reinforces implication #1. Solana’s composability drastically increases the capacity of developers to produce truly novel applications. Once produced, forking them to non-Solana chains is a tedious process.

*Neon Labs is bringing EVM compatibility to Solana after raising $40M in November. The alpha devnet was deployed on April 5th; beta mainnet is expected to launch in Q2. This will bring a substantial wave of Ethereum protocols and developers to Solana.

Due to its monolithic composability, high throughput, low transaction cost, and fork-resistant ecosystem, I expect Solana to outperform its existing competitors over the next 3-4 years. Occasional periods of performance degradation are concerning, but the Solana core developers are transparent and continually improving the underlying code issues causing these degradation periods. With over $2B in investment into the ecosystem over the last two years, a magnitude increase in users, developers, protocols, and market capitalization is realistic across the next four years.

Beyond the 3-4 year mark, assessing the growth and dominance of Solana becomes far more challenging. To become the global settlement layer, Solana will need to handle millions of transactions per second and eventually tens to hundreds of millions of transactions per second. The demand for throughput may increase at a rate exceeding Moore’s Law, or Moore’s Law models may become dislocated from the reality of Solana’s performance increases. The space is rapidly developing and four years is sufficient time for another Solana-esque protocol to completely upend the industry overnight.

While modular blockchains may face composability issues today, thousands of developers are working to solve composability degradation across roll-ups and shards. I have high confidence that those issues will be resolved in four years time, negating any composability advantage currently held by Solana. Additionally, Solana currently lacks a convincingly clear path to achieving nation-state level censorship resistance. On a high-enough timeframe, tail risks associated with nation-state attacks on a blockchain seeking to become the global settlement layer are significant. Hardware requirements are significant and, because all validator voting occurs on chain, the voting cost for a validator is roughly 1 SOL/day (~$35,000/year). This requires a validator to own a minimum of 4,600 SOL (~$460,000) in order to break even (excluding hardware costs). This is an exorbitantly high cost to individually validate the network and is prohibitive to decentralization.

This article was originally published on April 22nd, 2022. It was republished here without modification after the wallet address for shuel.eth was changed.

“The Web as I envisaged it, we have not seen it yet. The future is still so much bigger than the past.” -Tim Berners-Lee

The market is staggeringly saturated with smart contract platforms attempting to address the blockchain trilemma. 29 smart contract platforms currently command a market capitalization greater than $1,000,000,000. Dozens upon dozens of additional platforms have failed to achieve or receded below the $1B market capitalization line. During the next 3-4 years, I expect Solana to outperform all other existing non-Ethereum smart contract platforms. Key metrics by which to quantify this outperformance include, but are not limited to: daily active account growth %, market capitalization % increase, and total value locked (TVL) growth %. Solana will outperform its competitors because its design is distinctly different from the eerily similar design choices overlapping most of the market. Let’s examine what separates Solana from the rest of the market in detail.

Nearly all competitive smart contract platforms share a design choice: modularity. Modularity is generally defined as the decoupling of one or more of the consensus, execution, and data availability layers. The modular approach attempts to solve scalability bottlenecks that have traditionally plagued Ethereum. A review of the three layers and their respective bottlenecks can be found here. Modularity is juxtaposed with monolithic design, in which the three layers remain co-located. The general definition for modularity isn’t quite broad enough, however. A few examples will demonstrate why:

Example 1: Ethereum’s roll-ups are modular because they only perform execution; consensus and data availability are retained by the main Ethereum chain. The main Ethereum chain, however, still handles execution, consensus, and data-availability. In this sense, Ethereum is monolithic and only its roll-ups are modular. Polkadot employs a slightly varied design: parachains handle execution, while the relay chain handles consensus and data availability. The technical design differences are nuanced, complex, and outside the scope of this article.

Example 2: The Cosmos blockchain attempts to achieve scalability by acting as a “blockchain of blockchains” in which many independent blockchains are interconnected via the inter-blockchain communication (IBC) protocol. In this design, scalability is achieved by building a network of fragmented, monolithic blockchains that each handle their own consensus, execution, and data availability. Avalanche’s subnets are of a similar overall design: a network of fragmented, monolithic blockchains. This design is commonly referenced as being modular in nature.

Example 3: Ethereum 2.0, Zilliqa, and Near all employ sharding techniques, in which the data layer is spread out horizontally in order to improve scalability. This technique is modular in nature, as data becomes separated from consensus and execution.

Given these examples, modularity’s common definition is far too narrow. It may be more appropriately defined as any scaling solution that is not completely monolithic i.e. any design in which execution, data, and consensus for the entire network are not all handled in the same place.

Modularity improves scaling, but replaces it with another problem: composability is hindered. Composability is the capacity for system components to be recombined into larger structures and for the output of one component to be the input of another. This is commonly referred to as “money legos.”

Multicoin Capital’s Kyle Samani provides a succinct explanation of composability and its impacts on DeFi below:

Composability enables the development of higher order decentralized finance applications that have never previously existed. For example, the UXD protocol is a novel approach to algorithmic stablecoins using Solana DEXs (Mango currently) to fully collateralize issued stablecoins 1:1 with a delta-neutral position using derivatives. These protocols are easily composable on Solana because data, execution, and consensus for both happens in the same place.

Modular designs hinder composability for several reasons:

1. Syntactic composability is the ability of dApps to integrate and interact with one another. In avalanche, as an example, it is hindered (a) between the main network and its subnets and (b) subnet to subnet. Imagine that the UXD protocol exists on one subnet and Mango exists on another subnet. Every protocol interaction must be communicated across the two subnets. Now imagine a not-yet-developed protocol that requires interaction with a dozen other protocols spread across a dozen different subnets in order to generate its desired output. Bridging messages and tokens in that scenario would utilize excessive time that is unacceptable for many financial applications and increases failure risk. While Avalanche is developing solutions for rapid bridging and communications between subnets, monolithic designs will continue to retain a significant composability edge for the foreseeable future.

2. Atomic composability, the ability to bundle multiple operations into a single transaction, is impacted in a similar manner by modularization. Assets may become stranded or lost if one operation in the transaction succeeds while another fails. This is a risk commonly discussed regarding ETH 2.0’s sharding. Solana retains complete atomic composability due to its monolithic design.

No other non-Ethereum blockchain seriously competing to become a global settlement layer employs a fully monolithic design. This unique property will serve Solana well over the next several years. Solana developers will continue to produce novel DeFi products like UXD while competitor blockchains struggle to develop complex solutions to their composability issues.

The potential of composability cannot be overstated. Some of the coming financial creations will be truly revolutionary. Solana has over 300 DeFi applications currently running or under development. As developers create novel financial products, it will attract more developers, users, and TVL. This will result in a positive-feedback loop as the network’s growth leads to the incubation of additional applications and infrastructure on Solana.

Solana currently executes slightly under 3,000 transactions per second (tps) with block times of 400 milliseconds. It boasts a theoretical capacity of 65,000 tps. Furthermore, these transactions cost fractions of a penny - an average of $0.00025 / transaction. Founder Anatoly Yakovenko claims eight distinct innovations make this throughput and cost/tx possible. You can read more about the technical side of Solana here. Sub-second block times are necessary to attract traditional financial institutions that require rapid settlement times that aren’t tenable on chains like Ethereum.

Solana still advertises its mainnet as a beta product, despite its $7.36B TVL. Its reliability reflects this as well, with less-than-rare periods of severely degraded performance and infrequent complete outages. On September 14th 2021, the network went down completely for 17 hours during a denial-of-service attack.

https://twitter.com/SBF_FTX/status/1485839807708385283

Investors and Solana core developers take a collectively unconcerned approach to periods of degraded performance because (1) Solana still performs more TPS than most other blockchains while degraded and (2) it’s still advertised as a beta product under development. The developers are transparent and responsive in pushing code updates to the validators after each incident.

When compared against other Layer 1s, Solana appears most ready for developers to build applications for the masses despite its periods of degraded performance. There is no scaling solution coming at some arbitrary point in the future. The protocol is (mostly) ready to scale today.

Along with its monolithic design, Solana’s EVM non-compatibility is fairly unique. This has significant implications on the retention of resources within Solana’s network.

In EVM-compatible networks like Avalanche, protocols from Ethereum can easily be forked and deployed. Developers quickly fork previously successful protocols onto a new blockchain and collect available grant/incentive funds from that blockchain. These funds are then used to incentivize users to migrate. This historically produces cycles of mercenary capital rotation and vampire attacks, as investors and developers flow to the newest EVM-compatible blockchain with the best incentives. They then quickly move on to the next nearly-identical ecosystem once available incentives have been drained or a new blockchain offers slightly better subsidized incentives.

Solana utilizes the Rust programming language, which is substantially different from Solidity. This has several implications on retention of capital, users, and protocols:

1. Unique, successful applications built on Solana can’t easily be forked onto other blockchains. If individuals want to utilize the application, they must on-board into the Solana ecosystem. Bridging between Solana and EVM chains was originally difficult, further increasing the retention of capital and users on Solana. Bridging to and from Solana is rapidly becoming simplified and no longer poses a significant challenge to users, though.

2. Developers who dedicate time to building proficiency in Rust on Solana are far less likely to leave the ecosystem vs. Solidity developers who can jump to the next EVM chain with better incentives. With over $2,000,000,000 invested into the Solana ecosystem, I expect a magnitude increase in the number of protocols, developers, and users on Solana in 3-4 years. Those developers will be far more likely to stay working on Solana long-term compared to their peers in other ecosystems.

3. The previous discussion on composability reinforces implication #1. Solana’s composability drastically increases the capacity of developers to produce truly novel applications. Once produced, forking them to non-Solana chains is a tedious process.

*Neon Labs is bringing EVM compatibility to Solana after raising $40M in November. The alpha devnet was deployed on April 5th; beta mainnet is expected to launch in Q2. This will bring a substantial wave of Ethereum protocols and developers to Solana.

Due to its monolithic composability, high throughput, low transaction cost, and fork-resistant ecosystem, I expect Solana to outperform its existing competitors over the next 3-4 years. Occasional periods of performance degradation are concerning, but the Solana core developers are transparent and continually improving the underlying code issues causing these degradation periods. With over $2B in investment into the ecosystem over the last two years, a magnitude increase in users, developers, protocols, and market capitalization is realistic across the next four years.

Beyond the 3-4 year mark, assessing the growth and dominance of Solana becomes far more challenging. To become the global settlement layer, Solana will need to handle millions of transactions per second and eventually tens to hundreds of millions of transactions per second. The demand for throughput may increase at a rate exceeding Moore’s Law, or Moore’s Law models may become dislocated from the reality of Solana’s performance increases. The space is rapidly developing and four years is sufficient time for another Solana-esque protocol to completely upend the industry overnight.

While modular blockchains may face composability issues today, thousands of developers are working to solve composability degradation across roll-ups and shards. I have high confidence that those issues will be resolved in four years time, negating any composability advantage currently held by Solana. Additionally, Solana currently lacks a convincingly clear path to achieving nation-state level censorship resistance. On a high-enough timeframe, tail risks associated with nation-state attacks on a blockchain seeking to become the global settlement layer are significant. Hardware requirements are significant and, because all validator voting occurs on chain, the voting cost for a validator is roughly 1 SOL/day (~$35,000/year). This requires a validator to own a minimum of 4,600 SOL (~$460,000) in order to break even (excluding hardware costs). This is an exorbitantly high cost to individually validate the network and is prohibitive to decentralization.